Investigators will analyze medical students selection of graduate medical education (GME) programs (specialty and hospital location) in three cities where AIDS is concentrated: New York San Francisco and Los Angeles. Using data for 1983-1988. treads will be observed in each city In the paroutages of students choosing programs with high as compared to low AIDS preval (measured bv the number of AIDS patients In hospitals where program are located). and in choosing specialties in whlch AIDS patient care and exposure to potentially infective body fluids are likely (Internal Nedicine Surgery, Pediatrics). compared to other specialties. The study will assess the role in GME progran selection of student background factors. exposure to AIDS in medical school. student quality. and certain program factors using multivariate regression analysis. Analvse will be coducted at the individual level (graduating sedical students) and at the lustltutional level (prograns). Thus. investigators also will contrast characteristics of. high and low-AIDS GME programs, including thei size proportion of filled positions affillation arrangements and so on. Investigators also will compare the quality of students (using MCAT scores and other seasures) in programs with high and low numbers of AIDS patients. This project requires the creation of the AIDS and Residency Decision makin Database which will include information on GNE prograns between 1983 and 1988 as well as individual data on all neve entrants into GNE in the three study cities during the study period. These data are avallable in the Student and Applicant Information Management Systen (SAINS) of the Association of American Medical Colleges (AAMC). SAINS includes data on graduating student background factors, sedical school of origin and residency matched to. SAINS data will be linked to meatures of the prevalence of AIDS in residency hospitals, available from Hospital Councils in the three study cities. This database will support analysis of related insues in the future.